arXiv AI
What Does a ProcGen Generalization Gap Measure? Action Rules, Residual Entropy, and the Missing Random Floor
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Machine Learning
Aug 26
Where Entropy Is Measured Matters: Policy Geometry in Bounded Continuous-Control PPO
arXiv:2608.24488v1 Announce Type: new Abstract: Many continuous-control policies are optimized as unbounded Gaussians and then mapped into bounded actions. We show that where entropy is measured chan...
arXiv Machine Learning
Aug 28
Shared Actors Need Not Share Critics: Effects of Value Mismatch in Parallel Reinforcement Learning
The paper investigates the problem of sharing a single critic across multiple parallel environments in reinforcement learning. It shows that when environments assign different expected returns to the same state, a shared critic must reconcile conflicting value targets, which can distort advantage estimates and misguide policy updates. The authors propose a simple fix—providing the critic with the environment index—demonstrating through bandit models and experiments on CartPole, MuJoCo, BipedalWalker, and 16 Procgen games that this conditional critic stabilizes learning and boosts returns, achieving a 40.8% improvement in aggregate normalized return on unseen levels.
arXiv AI
Jul 3
AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents
arXiv:2607. 02255v1 Announce Type: new Abstract: Memory for a long-horizon LLM agent is a contract about what each future decision is allowed to see.
arXiv Machine Learning
Oct 7
What pass@k Cannot Measure: Evaluating Diversity and Capability Retention after Post-Training
The paper critiques the use of pass@$k$ as a sole metric for evaluating reinforcement‑learning post‑training on language models, showing that it ignores diversity in model outputs. Experiments on Qwen2.5‑1.5B‑Instruct trained with Group Relative Policy Optimization (GRPO) and rejection‑sampling fine‑tuning (RFT) reveal that these methods shift token‑level, answer‑level, and unique‑answer diversity in opposite directions, yet pass@$k$ fails to capture these differences. Even when restricting to verifier‑correct completions, diversity gaps persist, and pass@$k$ does not consistently identify the better performing arm on hard problem subsets, indicating that the metric can mask true capability changes.
arXiv AI
Oct 1
Free Everywhere, Exact on Trees: PPO's Dropped Correction Buys Sample Efficiency Under Aggressive Reuse
arXiv:2609.39634v1 Announce Type: cross Abstract: Common policy improvement methods, including TRPO, PPO, and GRPO, estimate policy improvement under the behavioral policy's state-visitation distribu...
arXiv Machine Learning
Jul 8
More Convincing, Not More Correct: Self-Play Reward Hacking of Reference-Free LLM Judges
arXiv:2607. 05904v1 Announce Type: new Abstract: Training a language model against its own reference-free judgments (the premise of self-rewarding, self-play, and LLM-as-a-judge pipelines) assumes a model's verdict on a shown answer tracks correctness.